Machine Translation on Tatoeba (it->en) benchmark leaderboard · SOTA2 Research

Machine Translation on Tatoeba (it->en)

79.0409sacreBLEUPhiMaestra-3

Evaluation Results

Method	Links
PhiMaestra-32026.01		79.0409	86.8107	0.8239	1.4275	0.9563	2.3486	0.787
Gemma-9B2026.01		72.0025	82.8853	0.7622	1.5577	0.9488	2.2472	0.7942
Gemma-9B2026.01		71.5484	82.5613	0.7585	1.5759	0.9483	2.2577	0.7941
Gemma-2B2026.01		70.8165	81.818	0.7313	1.6857	0.945	2.2465	0.7939
opus-mt-big2026.01		70.7301	81.7362	0.7357	1.7011	0.9437	2.3328	0.7878
opus-mt-big2026.01		70.7198	81.7337	0.7357	1.7016	0.9437	2.3341	0.7877
Madlad-7B2026.01		70.2099	82.1346	0.7559	1.593	0.9467	2.2606	0.7939
Gemma-2B2026.01		70.1194	81.206	0.7247	1.7165	0.9439	2.2672	0.7928
Madlad-7B2026.01		70.0311	81.7444	0.7493	1.6176	0.9457	2.283	0.7929
DIETA+cont2026.01		70.022	81.1897	0.7271	1.7418	0.9433	2.3449	0.7873
Madlad-3B2026.01		69.9979	81.6697	0.7417	1.6434	0.9447	2.266	0.7933
NLLB-1.3B2026.01		69.7703	80.6908	0.7281	1.6681	0.9415	2.3247	0.7875
DIETA+allsynth2026.01		69.7578	81.0969	0.724	1.7567	0.9433	2.3405	0.787
NLLB-3.3B2026.01		69.6032	80.6292	0.7251	1.6902	0.9411	2.3581	0.7851
opus-mt2026.01		69.6026	80.8182	0.7355	1.7269	0.9422	2.3936	0.7838
DIETA+nosynth2026.01		69.575	81.129	0.7285	1.7441	0.9412	2.3788	0.7836
Madlad-3B2026.01		69.5378	81.1643	0.7333	1.6818	0.9435	2.2939	0.7918
DIETA2026.01		69.0427	80.1357	0.727	1.6595	0.9386	2.3386	0.7818
NLLB-1.3B2026.01		68.9453	80.0527	0.7211	1.7102	0.9403	2.3626	0.7852
Tower-7B2026.01		68.7636	80.911	0.7223	1.7131	0.9434	2.2688	0.7929
NLLB-600M2026.01		67.6539	79.2751	0.6988	1.8047	0.9362	2.3621	0.7849
DIETA+BT2026.01		66.6445	78.9965	0.683	1.9522	0.9359	2.4627	0.7807
NLLB-600M2026.01		66.4912	78.4425	0.6849	1.8784	0.9337	2.4185	0.7818
Maestrale-v0.42026.01		59.0956	74.7694	0.633	1.9678	0.9281	2.4207	0.7788
mBART2026.01		58.8334	73.0329	0.5518	2.6164	0.9096	2.6754	0.7652
mBART2026.01		58.8334	73.0329	0.5518	2.6164	0.9096	2.6754	0.7652
LLaMAntino-8B2026.01		56.1724	75.2918	0.6205	2.0835	0.9243	2.4997	0.7734
EuroLLM-9B2026.01		55.0974	73.2035	0.6233	2.0345	0.9273	2.4353	0.7773
ModelloItalia-9B2026.01		51.6193	68.9674	0.421	2.9331	0.9014	2.9577	0.7348
Cerbero-7B2026.01		49.1672	62.58	0.1019	3.8588	0.8986	4.0801	0.6377
EuroLLM-1.7B2026.01		47.0759	67.7455	0.4588	2.787	0.8972	2.8982	0.7452
Minerva-7B2026.01		35.8318	55.4475	-0.5051	7.432	0.7723	7.7891	0.5322
Llama-3.1-8B2026.01		27.9916	39.7726	-0.773	6.516	0.8556	6.6222	0.4021